which X driver for NVIDIA Quadro FX 570M?

Gary Aitken freebsd at dreamchaser.org
Mon Feb 4 00:48:22 UTC 2013

On 02/03/13 16:11, Warren Block wrote:
> On Sun, 3 Feb 2013, Gary Aitken wrote:
>> For the OP, I (just now) got the nvidia driver and x to compile and run
>> under 9.1 as follows:
>> 1. make and install the x11/xorg port
>> 2. create the symlink /usr/X11R6 to point to /usr/local
>>   if /usr/X11R6 already exists as a regular directory and has stuff in it,
>>   move it aside or delete it.  I found mine only had stuff from previous
>>   attempts to build the nvidia driver in it (a "bin" plain file(!), and a
>>   "lib" containing files from the nvidia driver build), so I deleted it.
>> 3. download and unpack the nvidia driver;
>>   I used
>>     NVIDIA-FreeBSD-x86_64-304.64.tar.gz
>>   but note that that file is specific to my architecture, which is amd64;
>>   you'll need the appropriate one for your architecture.
>> 4. make the nvidia driver
>> 5. make sure step #4 added the line
>>     nvidia_load="YES"
>>   to /boot/loader.conf
>> 6. add the following lines to /etc/rc.conf if not already present
>>     # Enable hald and dbus for X to work with kbd and mouse
>>     hald_enable="YES"
>>     dbus_enable="YES"
> Why not just use the port?

When I tried that I got some message (can't remember what it was) saying the
port was obsolete or something like that implying it shouldn't be used.  I
thought the msg was in one of the posts I made but I can't find it.  I just
looked at the Makefile for the port and it looks like it is using the same
version of the driver so maybe it will work fine.  The message may also
have been the result of the port build being messed up because of the 
extraneous /usr/X11R6 directory I had at the time.
If I end up rebuilding I will try it.

More information about the freebsd-questions mailing list